Chem 481 Lecture Material 3/20/09 Radiation Detection and Measurement Semiconductor Detectors The electrons in a sample of silicon are each bound to specific silicon atoms (occupy the valence band). If sufficient energy is added to an electron, it is promoted to a band of overlapping energy levels that are not associated with specific atoms (conduction band) and is able to move freely within the solid lattice (see figure below). For semiconductors the size of the energy gap between the valence band and conduction band is in between that found in conductors (such as metals) and insulators. When a Group IVA semiconductor is doped with a Group VA element there will be an extra electron associated with the impurity atom. This results in an n-type semiconductor (see figure below) with an excess of negative charge carriers (electrons) and a donor energy level much closer to the conduction band is formed (see figure above). If a Group IIIA element is added, there is a vacancy (hole) created because of the fewer valence electrons associated with the impurity atom. This results in an p-type semiconductor (see figure below) with an excess of positive charge carriers (holes) and an acceptor energy level much closer to the valence band is formed (see figure above).
page 2 A semiconductor detector operates on the same principles as a reversed bias semiconductor diode. A diode consists of a junction between p-type and n-type semiconductor materials with an applied voltage such that the n-type is positive with respect to the p-type (see figure below). Electrons are drawn away from the junction in the n-type and the holes are drawn away from the junction in the p-type creating a depletion region with no excess charge carriers. Normally no current flows across this reversed-bias junction. If sufficient energy is added to this depletion region, electron-hole pairs will form analogous to the formation of ion pairs in a gas-filled detector. These charge carriers will then be collected by the electric field established by the reverse bias. For gamma rays to be detected efficiently, the depletion region must be much larger than can be created by increasing the applied voltage in a typical diode (2 mm) before the semiconductor breaks down. This is achieved by using high-purity (intrinsic) semiconductor material sandwiched between the n-type and p-type. To enhance the likelihood of a gamma ray interacting by the photoelectric effect (%Z 5 ), high-purity germanium (HPGe) is used instead of silicon.
page 3 Many different structural arrangements of the semiconductor materials are employed in HPGe detectors, however, the coaxial geometry shown below provides a detector with high efficiency and high resolution. The semiconductor material is housed inside an aluminum can. The extended range detector that we have in the lab (see figure above) has a thin beryllium window that extends the useful energy range of the detector down to a few kev (see figure below).
page 4 The band gap in germanium is only 0.74 ev and it takes only 2.98 ev to create an e - - hole pair in germanium ( corresponding values for Si are 1.12 ev and 3.61 ev). Consequently, to minimize the noise from thermal e - -hole pair formation, germanium (and Si) detectors must be cooled before use. This is achieved by mounting the detector in a vacuum chamber attached to a dipstick that is placed in a cryostat containing liquid nitrogen (see figure below). The output signal from a semiconductor detector is proportional to the energy deposited in the semiconductor; more energy deposited produces more e - -hole pairs and thus more charge is collected. Semiconductor detectors have superior energy resolution compared to NaI(Tl) detectors (see figure below) or proportional counters because of the small amount of energy required to produce the charge carriers that result in the output signal. For example, if 0.3 MeV of gamma energy is deposited, there will be approximately 10 5 e - -hole pairs formed in a germanium detector, 10 3 photoelectrons produced in a NaI(Tl) detector and 10 4 ion pairs produced in a proportional counter. This means that small fluctuations in the number of the charge carriers produced will be less significant for germanium detectors compared to NaI(Tl) detectors or proportional counters and that the output signal will be more reproducible. This results in improved energy resolution. The gas amplification that occurs in a proportional counter, or the electron multiplication that takes place in a NaI(Tl) detector, does nothing to improve the fundamental statistics of charge production in these devices.
page 5 The resolution of a germanium detector is usually expressed as the width at half height (FWHM) of the full-energy signal produced by the 1332-keV gamma ray emitted by 60 Co (see figure above). For a coaxial detector, such as the one in the lab, this FWHM is typically 1.8-2.0 kev. For a 3"x 3" NaI(Tl) detector the FWHM at 1332 kev is typically 60 kev. The counting efficiency of germanium detectors is usually expressed relative to the fullenergy counting efficiency of the 1332-keV gamma ray from 60 Co for a 3" x 3" cylindrical NaI(Tl) detector. The efficiency of the detector in the lab is 25%. Higher efficiencies require larger crystals of germanium. Detectors having a relative efficiency greater than 100% have been fabricated and require about 2 kg of high-purity germanium. Detectors based on silicon are useful for x-ray and alpha particle detection. X-ray detectors usually have a depletion region formed by drifting a donor impurity (lithium) into p-type silicon at elevated temperature. The lithium-drifted semiconductor is referred to a Si(Li) (pronounced silly) detector. Unlike lithium-drifted germanium detectors, Ge(Li) (pronounced jelly), which must be constantly cooled to liquid nitrogen temperature or the Li will diffuse out of the crystal, Si(Li) detectors need to be cooled only when in use. Charged particle detectors are used for alpha (and beta) particle detection. They typically consist of a very thin entrance window into a thin depletion region in n-type silicon. A surface barrier detector (see figure below) consists of n-type silicon that has been oxidized to produce a very thin layer of p-type material on the surface. A very thin evaporated gold contact is deposited on the oxide layer and an aluminum electrical contact is evaporated onto the n-type silicon. The gold contact and oxidized layer are much thinner than the range of alpha particles. A more rugged version of this is a PIPS (Passivated Implanted Planar Silicon) detector which has an implanted contact in place
page 6 of the surface contact and an entrance window typically < 50 nm. Charged particle detectors are usually operated at room temperature. Alpha detection is usually done by placing the detector in an vacuum chamber to minimize absorption of the alpha radiation before striking the detector.